Anthropic released Claude Sonnet 5, described as its most agentic Sonnet model yet, capable of planning, driving browsers and terminals, and running autonomously across long tasks. It is the default model for Free and Pro plans and is also live in Claude Code and the Claude Platform. Sonnet 5 beats Sonnet 4.6 on every published benchmark, scoring 63.2% on SWE-bench Pro, 81.2% on OSWorld-Verified, and 57.4% on HLE. Its intro pricing is $2/$10 per MTok through Aug 31, then $3/$15, while Opus 4.8 is $5/$25. Sonnet 5 uses an updated tokenizer, the same one introduced with Opus 4.7, where the same text can map to roughly 1.0 to 1.35 times more tokens. The model exposes effort levels: low, medium, high, and xhigh (extra high). At xhigh effort, it can cost more than Opus 4.8 for similar quality. Sonnet 5 is safer than 4.6 with deliberately low cyber capability, while Opus remains the pick for accuracy-critical work. On knowledge work (GDPval-AA v2), Sonnet 5 scores 1,618 and edges Opus 4.8’s 1,615.